source: app/src/gue/main.cc @ e2bde98

#include <time.h>
#include <stdlib.h>
#include "../lib/iotools.h"
#define _CMDLINE_FILE "../gue/cmdline.h"
#define CONFIGFILE1 "/home/ynka/utt/utt-0.9/conf/gue.conf"
#define CONFIGFILE2 "/home/ynka/utt/utt-0.9/conf/gue.conf"
#include "../common/common.h"
#include "common_guess.h"
#include "guess.h"
#include "cmdline.h"

#define W_SUFF 0.6
#define W_PREF 0.4

int main(int argc, char** argv) {

  int non_standard_config=0;

  gengetopt_args_info args;

  if(cmdline_parser(argc, argv, &args) != 0)
    exit(1);

  process_config_files(&args,argv[0]);

  process_common_options(&args,argv[0]);
  process_guess_options(&args);


  // PONIŻEJ POPRZEDNI KOD (JUSTYNY)
// //preliminary command-line parsing - for configuration file info only
//   gengetopt_args_info pre_args;

//   if (cmdline_parser(argc, argv, &pre_args) != 0)
//      exit(1);
//   if(pre_args.config_given){
//      printf("podano config: %s\n",pre_args.config_arg);
//      non_standard_config=1;
//   }
  

// //configuration file 1 parsing
//      struct cmdline_parser_params *params;   
//      params = cmdline_parser_params_init();
//      params->initialize = 1;
//      if(cmdline_parser_config_file(CONFIGFILE1,&args, params)!=0){
//              printf("System-wide configuration file parsing error!\n");              
//              exit(1);
//      }  

// //configuration file 2 parsing-overriding
//      params->initialize=0;
//      params->override=1;
//      char* config2=(non_standard_config)?pre_args.config_arg:CONFIGFILE2;
//      if(cmdline_parser_config_file(config2,&args, params)!=0){
//              printf("User configuration file parsing error!\n");             
//              return 1;
//      }

//      params->initialize=0;
//      params->override=1;
//      //params->check_required=1;
        
//      free(params);

// //command-line options parsing-overriding
//   if (cmdline_parser(argc, argv, &args) != 0)
//     exit(1);


  char line[MAX_LINE];
  char outline[MAX_LINE];
  char parms[MAX_LINE], desc[MAX_LINE], lemma[MAX_LINE];
  long line_count = 0;
  //  printf("d_f=%s\n", dict_file);
  Guess guess(dictionary);
  int words_count=0;
  time_t start_time = time(NULL);

  Segment seg;
  Words tab;
  char* form; //[MAX_FORM];
  while (fgets(line, MAX_LINE, inputf)==line) {
    line_count++;
    int start, len;

    line[strlen(line)-1] = '\0';

    if (!seg.parse(line)) {
      fprintf(stderr, "B³±d w wej¶ciu (linia: %d)\n", line_count);
      return -1;
    }

    if (process_seg(seg, args)) {
      words_count++;
      tab.clear();
      if (args.input_field_given>0) {
        form = getInput(args.input_field_arg, args.input_field_given, seg);
      } else
        form = seg.form;

      if (NULL == form) {
        continue;
      }

      guess.ana(form, tab);

      if ((tab.count()==0) && (!args.no_fail_flag)) {
        // no guesses - analysis was unsuccessful
        seg.print(outline); //this is necessary - seg.parse destroys line...
        fputs(outline, failedf);
        if (copy_processed)
          fputs(line, stdout);
        continue;
      }
      // we've got some guesses. Do we want to print it?
      if (args.only_fail_flag)
        continue;

      float last_weight=0;
      int i=0;
      int count=0;
      unsigned first=1;
      char* parms_end = parms;
      char last_lemma[MAX_LINE];

      while ((i=tab.next()) != -1 && count++<guess_count) {
        /* if we have "one-line" flag then everything goes in one segment as many fields,
         * if we have "one-field" flag everything goes in one segment as ONE field:
         * - diferent lemmas are separated with ';', sequent descriptions to one lemma
         *   are separated with ','
         */
        if ((!first) && (tab[i].w_suf() < cut_off) || (tab[i].w_suf() < delta * last_weight)) {
          break;
        }
        if (first) {
          parms_end += sprintf(parms_end, "%s", field_prefix);
        } else if (!args.one_field_flag)
          parms_end += sprintf(parms_end, "%s", field_prefix);

        if (!args.one_field_flag || strcmp(last_lemma, tab[i].lemma()) != 0) {
          if (args.one_field_flag && !first)
            parms_end += sprintf(parms_end, ";");
          parms_end += sprintf(parms_end, "%s", tab[i].lemma());
          strcpy(last_lemma, tab[i].lemma());
        }

        first=0;

        last_weight = tab[i].w_suf();
        if (!weights)
          parms_end += sprintf(parms_end, ",%s:%d",  tab[i].descr(), (int)tab[i].w_suf());
        else
          parms_end += sprintf(parms_end, ",%s", tab[i].descr());

        if (!args.one_field_flag) {
          seg.addfield(parms);
          parms_end = parms;
        }
        
        if (!(args.one_field_flag || args.one_line_flag)) {
          seg.print(outline);
          fputs(outline, outputf);
          --seg.auxn;
        }
        //if (copy_processed)
        //  fputs(outline, stdout);
      } //while

      if (args.one_field_flag)
        seg.addfield(parms);

      if (args.one_field_flag || args.one_line_flag){
        seg.print(outline);
        fputs(outline, outputf);
      }
    } else { // if (process_segment)
      // jak to nie jest wyraz - to przepisz token na wyjscie.
      //      printtok(line, start, len, cat, form);
      seg.print(outline);
      fputs(outline, outputf);
      if (copy_processed)
        fputs(outline, stdout);
    }
  }
  time_t end_time = time(NULL);
  if (per_info) {
    printf("Liczba s³ów: %d\n", words_count);
    printf("Czas analizy: %d sekund\n", end_time-start_time);
  }
  cmdline_parser_free(&args);
}